Distributed entrapment for multi-robot systems with uncertainties

In this paper we address the entrapment problem for a multi-robot system under the assumption of uncertainty in the knowledge of the target position. More precisely, we assume each robot models its knowledge of the location of the target through a Gaussian distribution, that is, with an expected value of the target location and the related covariance matrix. Motivated by this probabilistic modeling of the knowledge of the target location, we propose a novel algorithm where elliptical orbits are considered for the entrapment rather than circular ones, as in a classical entrapment formulation. A theoretical analysis of the entrapment algorithm properties is provided. In particular, we show this formulation to be a generalization of the classical entrapment scenarios. Simulation results are proposed to corroborate the theoretical analysis.